A vertical video is not a horizontal shot with the sides removed. The 9:16 frame changes where the subject can move, how the camera reveals space, and how much room remains for captions or interface overlays. Planning for the tall canvas before generation produces better short-form footage than cropping a wide result afterward.
Veo3o1 offers 9:16 as a direct aspect-ratio choice. It is available across the current creation workflows where supported, including Reference Images mode, which requires either 16:9 or 9:16 rather than Auto.
Compose for a tall frame
The center of a vertical frame is valuable. Keep the primary subject near the central column and avoid important details at the extreme top or bottom, where social interfaces may cover them.
Useful vertical compositions include:
- one person framed from the waist or chest up;
- a full-height product with controlled background space;
- an upward or downward camera move;
- a foreground-to-background reveal along the vertical axis;
- a close subject with negative space reserved for captions.
Crowded group scenes and wide lateral action are harder to read on a phone. If the action depends on a broad environment, consider 16:9 or break the idea into a close vertical shot and a separate establishing shot.
Say 9:16 ideas in the prompt
Selecting 9:16 sets the canvas, but the prompt should still describe a composition that belongs in that canvas.
Example:
A ceramic perfume bottle standing upright on a narrow stone pedestal, centered vertical composition with clear space above the cap for a headline. Soft mist rises behind it while the camera makes a slow upward push, controlled studio lighting, no text or logos added.
This prompt defines the subject, placement, useful negative space, motion, camera, and an exclusion. It does not rely on the aspect-ratio selector to solve composition by itself.
Design movement for mobile viewing
Short-form feeds move quickly, so the opening composition should be understandable without a long setup. Use one primary subject and one main action.
Good motion patterns for 9:16 include:
- a slow push toward a centered subject;
- a tilt from hands to face;
- an object rising into frame;
- a subject moving toward or away from the camera;
- a vertical reveal from foreground to background.
Use lateral pans carefully. Fast side-to-side motion can move the subject out of the narrow frame or make the result feel like a cropped landscape shot.
Use references without fighting the crop
If you use Reference Images, choose images that already leave enough room around the subject. A tightly cropped horizontal photo may not provide the model with the body, product shape, or background information needed for a convincing vertical composition.
For a recurring character, include at least one reference that shows the intended vertical framing. For a product, include a full-object view rather than only a close detail. Reference mode accepts one to three images and currently uses Lite or Fast with an 8-second duration.
Draft before final resolution
Start with 720p when testing framing and motion. Once the composition works, choose 1080p or 4K only if the destination and editing workflow need it. The model and resolution guide explains the current credit differences between 720p, 1080p, and 4K.
Review the result on an actual phone-sized viewport. Look for:
- a readable subject in the first moment;
- enough room for platform UI and captions;
- no important detail cut off at the edges;
- motion that stays inside the frame;
- a clean ending that can loop or cut naturally.
A repeatable 9:16 workflow
- Choose 9:16 before writing the final prompt.
- Define one subject, one action, and one camera behavior.
- Reserve intentional negative space if captions are planned.
- Use vertical-friendly references where identity matters.
- Draft at 720p and inspect on mobile.
- Refine one variable at a time.
- Render or enhance the selected shot at the delivery resolution.